1. Technical Field
The following embodiments generally relate to image processing. More particularly, an apparatus and method for capturing light field images are disclosed.
2. Description of the Related Art
Image capture apparatuses such as camera modules acquire color values that represent the image of a subject.
In order to acquire color values, an image capture apparatus includes an image sensor. The image sensor of the image capture apparatus has a width and a height, and has a predetermined resolution. The image sensor captures an image, and generates a color value for each of the coordinates of the image. Here, the coordinates indicate the position of a point on the image, and may be represented as (x, y), which is a value of 2D coordinates. Also, the color values, for example, may be an RGB value, which indicates the intensity of red, green, and blue.
Using such an image capture apparatus, a user may acquire a 2D color image. The existing image capture apparatus, which is referred to as a digital camera, generates a 2D color image as described above.
With the development of technology, a plenoptic technique is proposed. The plenoptic technique acquires both a color value for coordinates (x, y) of an image sensor and a color value for (u, v) at the said coordinates. Here, (u, v) may be information about angular resolution. In other words, the plenoptic technique acquires an (R, G, B) color value for (x, y, u, v).
According to the plenoptic technique, an array of micro-lenses is arranged between a main lens and an image sensor, and an (R, G, B) color value for (x, y, u, v) is acquired by the array of micro-lenses.
Using the plenoptic technique, information about the angle of light rays may be acquired. Based on this characteristic of the plenoptic technique, that is, because information about the angle of light rays may be acquired, the plenoptic technique is called a light field technique.
Existing image capture apparatuses such as camera modules acquire an image after focusing the apparatus on a subject. In contrast, a light field image capture apparatus using the light field technique acquires an image and may then perform optical refocusing for the acquired image. When images are acquired by existing image capture apparatuses, there may be images that are unsuitable for use or images that cannot be used because the subject is out of focus. However, the light field image capture apparatus may allow the focus to be freely readjusted. Therefore, when images are acquired by the light field image capture apparatus, there is no case in which the images cannot be used due to incorrect focusing. Also, the light field image apparatus may generate an all-in-focus image, which enables all the objects therein to be focused on.
The light field image capture apparatus also has a function of refocusing. The light field image capture apparatus records information about the directions in which light rays are travelling. Therefore, the light field image capture apparatus may generate an image from another viewpoint that is allowable with the recorded information about the light rays. Because images from various viewpoints may be generated, the light field image capture apparatus may be suitable for a multi-view stereoscopic display.
The light field image capture apparatus must partition the information space of a sensor respectively for (u, v) information, corresponding to angular resolution, and for (x, y) information, corresponding to spatial resolution. For example, when an image is captured at a high angular resolution, the spatial resolution of the image may be reduced. Conversely, when an image is captured at a high spatial resolution, the angular resolution of the image may be reduced. Therefore, when a light field image capture apparatus is designed, it is required to determine how to partition and store information about an angular resolution and information about a spatial resolution.
With regard to a light field image capture apparatus, Korean Patent Application Publication No. 10-2010-0088326 disclosed an apparatus of capturing multi-view light field data using a variable modulator.